Low discriminative power of WISC cognitive profile in developmental dyscalculia

Abstract

BackgroundThe role of domain-general cognitive abilities in the etiology of Developmental Dyscalculia (DD) is a hotly debated issue. AimsIn the present study, we tested whether WISC-IV cognitive profiles can be useful to single out DD. Methods and proceduresUsing a stringent 2-SD cutoff in a standardized numeracy battery, we identified children with DD (N = 43) within a clinical sample referred for assessment of learning disability and compared them in terms of WISC cognitive indexes to the remaining children without DD (N = 100) employing cross-validated logistic regression. Outcomes and resultsBoth groups showed higher Verbal Comprehension and Perceptual Reasoning than Working Memory and Processing Speed, and DD scores were generally lower. Predictive accuracy of WISC indexes in identifying DD individuals was low (AUC = 0.67) and it dropped to chance level in discriminating DD from selected controls (N = 43) with average math performance but matched on global IQ. The inclusion of a visuospatial memory score as an additional predictor did not improve classification accuracy. Conclusions and implicationsThese results demonstrate that cognitive profiles do not reliably discriminate DD from non-DD children, thereby weakening the appeal of domain-general accounts.